AI search for family law firms
Separation questions are asked privately and late, which makes an AI answer the first adviser many people meet. Joe Ashta works on those answers as a GEO / AEO consultant for law firms. This page covers what people ask, how the answers mix legal information with a few named local firms, why a respected practice can be missing from them, and why the individual lawyer is an entity of its own here.
What people ask, and how they ask it
The questions are emotionally loaded and procedurally uninformed. How separation actually works. Whether they have to leave the house. How custody and parenting time are decided. How support is calculated. Whether mediation is better than court, and what either costs. Whether they need a lawyer at all. People ask an AI system these things at eleven at night, before telling a friend, because the question is hard to say out loud. By the time a firm name is asked for, the system has the province, the stage of the separation and the issue in dispute.
How the answers are shaped
Family answers tend to run in two halves. The first is general legal information: how the process works in that jurisdiction, what the tests are, what the timelines are. The second names a few local firms or lawyers, usually with a reason each and a caution to get advice specific to the situation. That structure matters, because the first half is assembled from whichever pages explain the process best, and those pages are often published by law firms. A practice that explains parenting time clearly for its own province can be read into the first half and named in the second. A firm named in the text has a mention, and the source links beneath the answer are citations, pointing at pages read rather than firms named.
Why a well-regarded practice can be invisible
Family law runs on word of mouth more than almost any other consumer practice. Clients refer friends, therapists and financial advisers refer clients, and other lawyers refer conflicts out. A practice can be busy for a decade on that alone, with a small website, thin biographies and a stale directory listing. That reputation is real, and it is held in people's heads.
An AI system builds its picture from pages, so a practice with a thin public record presents as a thin candidate, whatever the local bar thinks of it. The firm every mediator in the city recommends can be absent from the answer given to the person those mediators would have sent, and the absence is silent. How the shortlist is built is set out on How AI chooses which law firms to name.
The lawyer is an entity too
In family law, buyers ask for a person as often as a firm, and answers name individual lawyers alongside firms. Each lawyer is therefore a separate entity the system has to resolve, and every inconsistency in the lawyer record weakens both the lawyer and the practice.
Three problems recur. Biographies that disagree with the law society record on call year, name form or practice focus. Lawyers who moved firms, leaving a live biography at the previous firm and press that still attaches them to it, so a system reading all of that may place the lawyer in two firms at once. And practices whose lawyers appear only on a shared team page, with no individual page to attach evidence to. The fixes are dull and effective: one biography page per lawyer, agreeing with the licensing record; the same name form everywhere; directory profiles kept current. Structured data connecting a lawyer to a firm, an office and a licence is how these records reach the reference sources described on Knowledge graphs, and becoming one clear, corroborated node is Entity authority.
Pages worth citing
Three kinds earn their place. Process explanations written for the jurisdiction the firm practises in, covering separation, divorce, parenting arrangements, support, property division, mediation and litigation. Jurisdiction-specific answers, because family law rules differ by province and state and a generic page is useless to a system answering a local question. And fee transparency: how the firm charges, what a consultation costs, what an uncontested matter runs to. Cost is asked early and published late, so a page answering it plainly is retrieved where competitors are silent.
What to measure and fix
List the questions a separating person asks in the firm's province, from the earliest process question to the request for a name. Establish which firms and lawyers the platforms name today, and whether the practice appears correctly and in the right city. Then fix the lawyer records, publish the missing process, fee and jurisdiction pages, and seek corroboration from sources the firm does not control. Report it as a mention rate, and treat a recommendation as the outcome that counts. Measuring that and fixing what it exposes for such a firm is what Joe Ashta does as a GEO / AEO consultant, and SEO and Google Ads engagements run through NearMe Marketing. The overview is Law firm AI search, the role is GEO consultant for law firms, and the canonical facts sit on the entity record.
Related: AI search for criminal defence firms, AI search for personal injury firms (coming soon) and AI search for BigLaw.